DESCRIPTION: adapted from applicant's abstract): Memory loss ranks high among the health issues that concern senior citizens. Most of the elderly affected have memory deficits that are not directly attributable to vascular or degenerative disorders. Hence, there is a great need to conceptualize and test other candidate mechanisms for these impairments. The goal of this proposal is to assess, among non-demented, non-diabetic elderly, the impact of poor peripheral glucose regulation and cortisol dysregulation on hippocampal volume and memory performance. Hippocampal structure will be assessed by means of MRI-derived volumes and its function by declarative memory assessments. The applicants predict that individuals with poor glucose regulation or cortisol dysregulation will have specific reductions in hippocampal volumes that will be associated with specific impairments in declarative memory performance. The applicants have strong preliminary data pointing to the feasibility of these hypothesized cross-sectional relationships. In addition, the hypothesize that there will be a cumulative effect; individuals with both poor glucose regulation and cortisol dysregulation will be more affected. The applicants also predict that longitudinal declines in glucose regulation or increases in cortisol dysregulation will place elderly individuals at risk for larger losses in hippocampal structure and function. Older women are more likely than men to have alterations in cortisol feedback inhibition, to have increased vulnerability to elevated cortisol levels, and to show stronger age-related hippocampal atrophy. Consequently, the impact of gender in the hypothesized relationships will specifically examine. The applicants will study individuals spanning the whole range of cognitive performance in the non-demented range. A 120 volunteers, 60 to 75 years of age and 50 percent female so as to have over 40 subjects in each gender group at follow-up, assuming a 30 percent attrition rate at the two-year follow-up will be assess. The applicants have developed valid and reliable MRI-based hippocampal measurement methods and have a long history of working with the population proposed for study.